Obesity disproportionately affects women, and contributes to their increasing rate of type 2 diabetes and cardiovascular disease. Although adiposity increases the risk of diabetes, the contribution of fitness to impaired glucose homeostasis, especially in an increasingly sedentary workforce, is unknown.We propose that exercise fitness may contribute to glucose homeostasis in overweight and obese women who have impaired insulin sensitivity. To determine the contribution of fitness, we measured adiposity, exercise fitness, insulin sensitivity and non-insulin-mediated glucose metabolism in sedentary overweight and obese women employed at the National Institutes of Health.Seventy non-diabetic women BMI, 33.36.2 kg/m2;age, 4510 years, (meanSD) were enrolled. Adiposity was assessed by waist circumference (WC) and by dual-energy X-ray absorptiometry (DXA) measurements of truncal fat and lean body mass (LBM). Exercise fitness was measured by peak oxygen consumption (VO2 peak) during graded treadmill exercise using the Bruce Protocol. To minimize the impact of excess adipose tissue in the obese, VO2 peak was normalized to LBM to better estimate the contribution of exercising skeletal muscle to O2 uptake. We also measured exercise duration, respiratory exchange ratio (RER: measure of the relative contribution of fat to carbohydrate as an energy source for exercising muscle) and anaerobic threshold (AT: index of lactate accumulation). Insulin sensitivity (SI) and glucose effectiveness (SG: measure of non-insulin-mediated glucose disposal) were determined by the minimal model. Multivariate analysis was performed to determine predictors of SI. SI, SG, exercise duration, AT and RER were compared by tertiles of VO2 peak.Multiple regression analysis for predicting SI (range: 0.49 to 13.90 L.U-1.min-1) using stepwise modeling with age (23 to 66 years), WC (77.2 to 145.7 cm), truncal fat (31.3 to 58.5%), and VO2 peak (25.9 to 72.1 ml/kg LBM/min) revealed truncal fat (R2= 0.18, P<0.001) and VO2 peak (R2= 0.25, P<0.02) as independent predictors of SI. Women in the lowest tertile of VO2 peak (34.53.8 ml/kg LBM/min) had significantly lower exercise duration (337107 vs. 48882 seconds), AT (1.40.2 vs. 1.60.3 L/min) and RER (1.050.1 vs. 1.150.1) (all P<0.02), compared with women in the highest tertile of VO2 peak (54.36.0 ml/kg LBM/min). Furthermore, the lowest tertile indicated significantly lower SG (0.010.006 vs. 0.020.007 min-1) and SI (2.451.26 vs. 4.652.49 L. U-1.min-1) compared the highest tertile (P<0.001).Conclusion Although adiposity is associated with diminished insulin sensitivity in sedentary overweight and obese women, poor fitness is also a major factor. In addition, reduced non-insulin-mediated glucose metabolism by skeletal muscle may contribute to impaired glucose homeostasis in poorly fit women. Recent studies suggest that some patients with, or at risk for, atherosclerosis may have dysfunctional high-density lipoprotein (HDL) despite normal cholesterol content. An HDL-associated enzymeparaoxonase 1 (PON1)protects lipoproteins from oxidation by degrading oxidized lipids including oxidized cholesteryl esters and phospholipids.We assessed the hypothesis that anti-oxidant properties of HDL might be abnormal due to reduced activity of PON1 in women at risk for atherosclerosis because of sedentary lifestyle, obesity, and diminished insulin sensitivity. Forty-nine non-diabetic overweight and obese women (23 Caucasian, 26 African American) were enrolled in a worksite wellness program at the National Institutes of Health. Adiposity was assessed by BMI and by waist circumference. PON1 activity was assessed by the rate of cleavage of phenyl acetate by arylesterase/paraoxonase, resulting in phenol formation measured by absorbance at 270 nm. Insulin sensitivity (SI) was determined by the minimal model. To determine whether PON1 activity might preserve HDL in an intact state, the Oxygen Radical Absorbance Capacity (ORAC) assay was used to test the susceptibility of HDL to oxidant stress, and was assessed by measuring the capacity of HDL isolated from serum to inhibit oxygen radicals produced by 2,2-azobis(2-amidino-propane) dihydrochloride every minute over 2 hours. Data were analyzed by comparing measurements from women in the lowest tertile of SI (more insulin resistant) to those from women in the highest tertile of SI (more insulin sensitive). Data are reported as mean values SD. Women in the lowest tertile of SI (SI<2.54 L.U-1.min-1) were similar in age to women in the highest tertile of SI (SI>4.00 L.mU-1.min-1) (44 13 vs. 44 11 years, P=0.91), but were significantly more obese (BMI 38.8 8.3 vs. 29.7 3.1 kg/m2, P<0.001 and waist circumference 115.5 20.0 vs. 99.2 11.5 cm, P=0.008). Women in the lowest tertile of SI had marginally lower HDL cholesterol levels compared with women in the highest tertile (54 18 vs. 64 17 mg/dL, P=0.12). PON1 activity was significantly reduced in women with the lowest tertile of SI compared with women in the highest tertile (120.2 19.8 vs. 144.6 29.7 kU/L, P=0.01). For the cohort, PON1 activity was significantly associated with resistance of HDL to oxidation as measured by ORAC normalized to HDL cholesterol content (r=0.39, P<0.01). We conclude that reduced HDL-associated PON1 activity may be associated with diabetes risk, as women who were more obese and insulin resistant had significantly lower enzyme activity than women who were less obese and more insulin sensitive. Furthermore, the correlation between PON1 activity and susceptibility of HDL to oxidation suggests that PON1 is an important determinant of overall HDL anti-oxidant capacity.